Plant Functional Trait Variation And Community Assembly

Exploring the structure, causes and consequences of functional trait variation in biological communities is crucial to understand community assembly. Based on intensive sampling of five key functional traits(leaf size, specific leaf area, leaf dry matter content, wood density and wood dry matter content) in a 50-ha subtropical forest(Heishiding), this thesis attempts to(i) quantify trait composition and structure of the study tree community,(ii) examine the interspecific and intraspecific trait variation in relation to habitat variation,(iii) map species niche positions, width and overlaps along trait spectrum and environmental gradients, and(iv) link interspecific trait similarity with spatial associations.In Chapter 2 I analyzed trait composition and structure of the tree community of the Heishiding plot. Results show that interspecific variation(accounting for 57-72% of total variance) overrides intraspecific variation in all five traits. This interspecific variation could be related to species sorting over long environmental gradients. Multivariate trait relationships are similar at the species and individual level, generally converging into a “plant economics spectrum” with low specific leaf area and high wood density at the one end(“slow return” syndrome) and the opposite syndrome at the other end, which suggests that the common evolutionary and biophysical constraints upon species strategies also work in individuals within species to some extent. However, the leading dimension of plant economics spectrum is less evident within species than across species. This suggests that individual trees can flexibly adjust their trait to cope with immediate environments, which can be uncovered only by focusing on intraspecific variation. Interspecific trait dispersion in both univariate and multivariate trait spectra is more pronounced than randomly expected, but trait evenness does not depart from random expectations, indicating divergent trait specialization between species or at least between functional groups can account for species coexistence of this forest community.In Chapter 3 I examined the interspecific and intraspecific trait variation along soil and light gradients. Light availability and total nitrogen, available aluminum in soil have significant relationships with trait variation at both interspecific and intraspecific levels, thereby comprising the dominant environmental gradients in the study forest. In general, habitats with higher soil aluminum, lower soil nitrogen and light availability are associated with species and individuals with lower specific leaf area, higher dry matter content and denser wood. This conforms to the theory of plant economics spectrum: harsh habitats select for “slow return” species. For most trait-habitat combinations, their relationships follow the same direction at both interspecific and intraspecific levels. But specific leaf area increases with decreasing light availability within species while species with lower specific leaf area are restricted to more shaded habitats, pointing to the importance of intraspecific variation in relation to proximate habitats. The magnitude of intraspecific variation of each species is also related to species-specific average habitats, with low-variability species restricted to more stressed habitats(i.e. higher soil aluminum, lower soil nitrogen and light availability), suggesting that intraspecific variability may be a consequence of habitat specialization and in turn can be a driving force of community assembly.In Chapter 4 I examined whether trait- and habitat-defined niches were concordantly structured and how they were correlated with species abundance. I found that while interspecific overlaps for both trait- and habitat-based niches were lower than expected by chance, species with common traits also inhabited common habitats and species pairs with higher trait overlap also shared more similar habitats. Furthermore, species with niche positions closer to community-wide trait- and habitat-defined niche modes tended to be dominant, and species pairs with higher trait and habitat overlaps tended to be co-dominant. Our results suggest the important role of species sorting along environmental gradients. However, no correlation was found between trait- and habitat-defined niche size, nor for their relationships with species abundance, indicating less importance of niche size than of niche positions in structuring this forest community. This study shows trait- and habitat-defined niches are consistent in measurement.In Chapter 5 I evaluated the relationships between interspecific trait similarity and spatial associations of tree species in the study forest. Out of all 11935 possible species pairs for 155 common species, 7766 pairs(65% in proportion) occurred as nearest neighbors and it is more frequent than expected by chance. However, rarer species occurred as nearest neighbors less frequently than more common species did. This suggests a role of stochastic processes in shaping species coexistence in neighborhood. Species pairs with high trait similarity have high probabilities of being nearest neighbors and short reciprocal minimum distances which indicates the importance of habitat filtering rather than competitive interaction. Altogether, interspecific trait similarity can be an important factor deterministically influencing species coexistence.